A reader asked:
A fellow trader told me “the trend is your friend and you should trade in the direction of the trend, unless there is a major divergence. If there is a major divergence, the divergence is your BETTER friend and probability favors trades (in the next lower time frame, ala Elder’s triple screen) in the direction of the major divergence.” Do you agree with this statement? Is this something you have studied? An example of a major indicator divergence is defined as a MACD line and histogram divergence, when the MACD hist rises/falls below 0.
This is a great question as it offers us a chance to argue the bigger picture. There is an assortment of indicators one could use for entry and or exit. For example, some of the best traders use price breakouts. But fixating on a magical indicator solution misses the point unfortunately. What should you do instead? Answer these 5 questions:
- How does the system determine what market to buy or sell at any time?
- How does the system determine how much of a market to buy or sell at any time?
- How does the system determine when you buy or sell a market?
- How does the system determine when you get out of a losing position?
- How does the system determine when you get out of a winning position?
Answer these 5 questions first. If your divergence indicator idea is part of answering the 5 questions, go for it, after sufficient testing.
Why the Five Questions Matter More Than Any Indicator
The divergence indicator question is a common form of a much broader question: can I use this specific signal to improve my trading? The answer to that question is always the same: only if it is integrated into a complete system that answers all five questions. A signal that tells you when to enter but not when to exit, how much to trade, or how to select markets is not a trading system. It is one component of a potentially useful system that does not yet exist.
The five questions define the complete anatomy of a trading system. Question 1 is market selection. Question 2 is position sizing. Questions 3, 4, and 5 are entry and exit criteria. A signal like MACD divergence can potentially contribute an answer to Question 3, but it says nothing about Questions 1, 2, 4, or 5. Without answers to those four questions, the entry signal cannot be evaluated as part of a system because the system does not exist.
The specific claim that divergence is a “better friend” than the trend is testable but needs to be tested in the context of a complete system. If MACD divergence signals consistently produce better entries than price breakout signals when all five questions are answered and the system is tested across a sufficient historical sample, then the claim has empirical support. If MACD divergence produces worse entries, or if the complexity of incorporating divergence into the system reduces its robustness without improving its edge, then the claim fails the test.
The warning about fixating on magical indicator solutions is the critical point. The trading seminar industry generates enormous revenue by selling the premise that there is a specific signal, indicator, or pattern that unlocks consistent profits. Each new indicator feels like it might be the one. The MACD divergence is compelling because it incorporates momentum, trend direction, and a measure of trend strength simultaneously. It feels sophisticated. But sophistication in an indicator does not translate to edge in a system. Edge comes from the system’s five components working together across a sufficient sample of trades in real market conditions.
The “after sufficient testing” qualifier is not a polite hedge. It is the most important part of the answer. An indicator that looks promising on a chart does not constitute sufficient testing. Sufficient testing means applying the complete system with the indicator as one component to a historical sample that spans multiple market regimes, measuring whether the system’s performance is robust to variation in the indicator’s parameters, and verifying that the performance reflects genuine market structure rather than historical data fitting.
Frequently Asked Questions
Why do the five questions matter more than indicator selection?
Because without answers to all five questions, there is no system to evaluate. An entry signal tells you when to enter. Without defined position sizing, you do not know how much to enter. Without defined exit criteria for both losses and wins, you do not know when to exit. The entry signal’s contribution to system performance cannot be measured in isolation because performance is a function of all five components interacting. Fixating on the entry signal while ignoring the other four components is building on an incomplete foundation.
Is MACD divergence a valid trend following entry signal?
It is a testable hypothesis that requires integration into a complete system before it can be evaluated. If MACD divergence, when combined with defined market selection criteria, position sizing rules, and exit criteria for both losses and wins, produces a system with positive expected value across a sufficient historical sample that is robust to parameter variation, then it is a valid component of that specific system. The validity is determined by the complete system’s performance, not by the indicator’s theoretical appeal.
What does Elder’s triple screen system do differently from simple divergence trading?
Elder’s triple screen uses multiple time frames to filter entries: the higher time frame defines the trend direction, the intermediate time frame identifies corrections within the trend, and the lower time frame provides the specific entry timing. Divergence in the intermediate time frame is the signal to look for entries in the lower time frame in the direction of the higher time frame trend. This multi-frame structure addresses Question 3 (when to enter) but still requires complete answers to the other four questions to constitute a tradeable system.
Want to learn more and start trading trend following systems?
Start here